import csv
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
import sklearn
import plotly.express as px
import plotly.figure_factory as ff
import plotly.offline as pyo
from IPython.display import Image
import math
from itables import init_notebook_mode

sns.set(rc={'figure.figsize':(11.7,8.27)})
pyo.init_notebook_mode()
init_notebook_mode(all_interactive=True)

Part 1: Examing the dataset

df = pd.read_csv('./dinosaur_cleaned.csv')

df

name	diet	period	lived_in	type	length	taxonomy	named_by	species	link
Loading... (need help?)

histograms = df.hist()

px.scatter(df, x = 'period', y = 'length', color = 'diet', hover_data = ['name'])

px.scatter(df, x = 'period', y = 'length', color = 'type', hover_data = ['name'])

px.scatter(df, x = 'period', y= 'length', trendline = 'ols', trendline_color_override = 'darkblue')

Image('google.png')

Part 2: Exploration

diet_vs_length_scatter = sns.catplot(x = 'diet', y = 'length', data = df, height = 8.27, aspect = 11.7/8.27)

diet_vs_length_box = sns.catplot(x = 'diet', y = 'length', data = df, kind= 'box', height = 8.27, aspect = 11.7/8.27)

type_vs_length_cat = sns.catplot(x = 'type', y = 'length', data = df, height = 8.27, aspect = 11.7/8.27)

type_vs_length_box = sns.catplot(x = 'type', y = 'length', data = df, kind = 'box', height = 8.27, aspect = 11.7/8.27)

Part 3: Research question

-> Are herbivores longer than carnivores on average?

μ1 -> herbivore mean

μ2 -> carnivore mean

H0: μ1 ≤ μ2

Ha: μ1 > μ2

Selected confidence interval = 95%

Image('formula.png')

df['diet'].value_counts()

	diet
Loading... (need help?)

herbivore_mean_x1 = df.loc[df['diet'] == 'herbivorous']['length'].mean()
herbivore_std_dev_s1 = df.loc[df['diet'] == 'herbivorous']['length'].std()
herbivore_count_n1 = df.loc[df['diet'] == 'herbivorous']['length'].count()


carnivore_mean_x2 = df.loc[df['diet'] == 'carnivorous']['length'].mean()
carnivore_std_dev_s2 = df.loc[df['diet'] == 'carnivorous']['length'].std()
carnivore_count_n2 = df.loc[df['diet'] == 'carnivorous']['length'].count()

difference_d0 = 0

print("OUTPUT:")
print("x1 = " + str(herbivore_mean_x1))
print("s1 = " + str(herbivore_std_dev_s1))
print("n1 = " + str(herbivore_count_n1))

print("x2 = " + str(carnivore_mean_x2))
print("s2 = " + str(carnivore_std_dev_s2))
print("n2 = " + str(carnivore_count_n2))

OUTPUT:
x1 = 8.899378881987579
s1 = 7.226483384869934
n1 = 161
x2 = 5.220222222222222
s2 = 3.9054982785804686
n2 = 90

numerator = (herbivore_mean_x1 - carnivore_mean_x2) - difference_d0
denominator = math.sqrt( herbivore_std_dev_s1 / herbivore_count_n1 + carnivore_std_dev_s2 / carnivore_count_n2)

z = numerator / denominator

print('z = ' + str(z))

z = 12.38279139344993

Since P(Z > 12.38) < 0.05

We reject the null.

This means that herbivores are on average larger than carnivores.

Part 4: For fun

year_list = []

for name in df['named_by']:
    if int(name[-5:-1]) > 1000:
        year_list.append(int(name[-5:-1]))
    else:
        year_list.append(2007)

df.insert(len(df.columns) - 1, 'year_of_discovery', year_list)

df

name	diet	period	lived_in	type	length	taxonomy	named_by	species	year_of_discovery	link
Loading... (need help?)

px.scatter_3d(df, x = 'period', y = 'length', z = 'year_of_discovery', hover_data = ['name'])